VLF AUTOMATIC GROUND BALANCE PRINCIPLES

[DOOLEY]

hello all.

I have a question for "the people in the know".


for a while now I have been wondering how automatic ground balance works.

manual ground balance I got my head around some time ago.

what are the principles of automatic ground balance ??
are there many ways of doing it ??

can it be done with analogue without using micro's ??
(sorry, i'm an analogue person, 4000' series is as digital as I go........lol. )

although micro examples welcome,
with block diagrams, logical rules.

what are the logical principle / rules behind how it all works ???

circuit examples to show how the hell it all works much appreciated.
as much information on this subject would be welcome.

been swinging away allot recently and just wondered "how the hell does it do that"

thank you..........

[Skippy]

There's some background reading here:
http://jb-ms.com/Baron/gb.htm

[DOOLEY]

cheers skippy,

I remember that article ,
read it a while ago,
interesting reading,
think it was posted on here.

unfortunately it didn't give much away as to how its all done.

although it does suggest that, although poss to do it purely with analogue
maybe not the best way of doing it.

would be nice to see the add on circuit that was used.........

[Davor]

The sure-fire way is multifrequency approach where you subtract contributions of the soil and what you get are only targets. It should work even on difficult terrains, but at a cost of somewhat reduced sensitivity (due to subtraction).
The way to auto-tune a single frequency VLF rig to any kind of balance, including ground balance, is to use a Costas loop. So for example, if you balance your rig in presence of a target, it will not see that target. For that reason any kind of self-adjusting approach may drive you crazy, as the targets start "disappearing" as you approach them. For that particular reason the interactive automatic GB procedure seem only to make sense.
On the other hand, the manual GB is not too different from an interactive automatic GB from the user's point of view, so why bother with automatic at all.

[Skippy]

The Bounty Hunter Tracker 4 has a form of auto-ground-balance that will adjust to wet salt, I understand. Very long time-constant are used, I'm sure, it's known to be slow in real life, and I assume that it's a simple way round the problem of 'tracking off' on actual buried targets. Maybe the circuit for that machine is available somewhere?

[DOOLEY]

multi frequency is another one of my "how does that all work" questions,
poss try and get my head round the principles of that at a later date.

yes, can imagine some of the more delicate signals could be quite quickly tracked out
and disappear.

poss a manual / automatic system,
so without fine tuning a knob very delicately,
which can be tricky esp if its been set up to go all the way to salt........

anyone have any schematics of systems that were in use on older machines ???

sort of thinking of push button to activate and track, then stay there, fixed..........

I know,
my request does sound a bit newbie .

[Davor]

What you suggest would be an interactive approach, and it seem to be the best, but again, it is not too different for a user from a manual GB.

[DOOLEY]

one of the things I was thinking was using a digi-pot,
say a WMS7100 in replacement of the GB POT.

and having it initially on up down buttons.

256 presses to go from end to end is allot of presses.
on a 50k digipot with 256 steps that's 190 ohm per step,
hopefully that would be enough resolution .

was more thinking of driving the digipot with a clock,
say 10-20hz.
then using a pair of comparators set up with a "window"
to alter the up down pins on the digipot until balance is found.

not sure how workable this will be,
going to take a bit of thinking about.........

[Davor]

It can work that way. Criterion for up or down may be decided from a quadrant ground signal is responding from a GB phase. Say, even quadrants: up, odd quadrants: down. Or even simpler, quadrants 3 and 4 do nothing, quadrant 1: down, quadrant 2: up. The 2nd option can easily be implemented, also with a threshold.

[Skippy]

One possible problem with the digital pot is the quantised nature of its steps. Each step could result in a false signal (or a silent null, depending on step direction) later in the signal processing chain.

[Davor]

Same goes for non-smooth potentiometers. With 256 steps, I think you are well covered.

[Skippy]

If you're just making a land machine, and the 256 steps only cover 0 to 10 degrees lag (relative to ferrite) then it will likely be OK, but if you plan on covering 90 degrees to wet salt, then there are possible problems.

[Davor]

I'd say 1/3 of a degree granularity is fine enough, but surely the transition is not continuous.

[DOOLEY]

so digipot is a poss then.

a switch and a pre-set multi turn for going to salt poss the best way,
as you say 0-10 in "normal position"

then salt position can extend the other way.

still pondering a way to do this automatically with analogue .
may be pondering for some time...........

[Skippy]

One 'unconventional' way to implement this in an analogue way is to use an LDR (light-dependant resistor, like CdS) and an illuminating LED. A more sophisticated method uses two 'identical' LDR's, one is used in a feedback loop to control linearity, the other should track it. Or for a more thorough job, there are 'all-in-one' FET-based devices:
http://www.edn.com/design/analog/436...-potentiometer

It depends on what the original circuit is like. There may be simpler ways of achieving the required function.